Viva la revolution!

Further my Vodafone lecture blog on LewyLand the other week i just realised the existence of Fon, and how it fits in quite snuggly with the ideas in that post; if it takes off as quickly as i think it might, then we could be looking at a speed of mobile data network that i thought was several years away due to the bandwidth-range trade off problem i was talking about.

Basically, Fon is an attempt to create a high speed wireless internet network using millions of short range WiFi hubs. But having private users and high street entrepreneurs own and run the equipment from their own private board band connections! It could well work, as they are subsidising their wireless routers (by selling them at a loss now while building up the network as fast as possible) making them as cheap or cheaper than standard ones u'd use to share you broad band anyway! (so y not get one guys?!)

It's very ambitious, mostly with respect to actually having the technology work reliably (WiFi ain't all that stable when u're sat at home by your router, let alone wondering the streets!). Getting applications like Skype to seamlessly hand over between hubs would be key (i presume that's not possible yet). Idealy it would auto fall back to a 3G (descendant) mobile service when out of Fon coverage. But Prof. Mike Walker assured us that would be too technically challenging...(well that would be the line of a mobile carrier with regards to a technology specifically designed to ease the transition way from their services!)

In theory it could kick mobile phone operators up the arse, hopefully spurring them to bring out better/cheaper mobile data services asap in preemptive competition. Ideally, one of the operators would/will realise the massive benefits of this technology movement and team up with Fon, forming a robust hybrid mobile data network. It would trash the cometition in cities and build up locations once the handsets/software becomes reliable enough to cope with the more complex system.

(Above: orange dots and green smudges are hub locations in Reading. Given that i've only just heard of Fon, that's not bad! But there will need to be a few doublings in popularity.)

The Fon network would be ideal for street level navigation applications, especially in cities where GPS type signals maybe particularly unreliable. U wouldn't need more than the faintest of WiFi signals to identify a hub, look it up in a POI (points of interest) type list and start narrowing down your position. It would be far easier to install WiFi hubs in underground locations (tube stations and maybe even road tunnels when handover techniques step up a level or 2. I forgot to check if the London Underground has any type of mobile phone service the other week) and on planes/ships/inside major buildings.... A Fon mobile carrier partnership would have good incentive to install infrastructure in busy public transport locations as it would give another edge over pure mobile Operations. I recon they'd have to start following suit with they're own WiFi/WiMax hardware if they couldn't get a similar deal with Fon (or an equally ubiquitous entity). Cool

[Imported from "LewyLife"] The sweet taste of hypomania!

So "Lewy's Life" pretty much equates to a running up date on my health now!:

Remember the diary free thing? Well i started a follow up of wheat free (just in case my negative blood test for Coeliac the other month wasn't up to scratch) which lasted til after breakfast when i nearly fell asleep (and I'd only had bacon, scrambled egg, mushroom and baked beans!).

With exams coming up (now mid way, and not doing at all bad considering), and after a whole bank holiday weekend of sleeping midnight til 10am but being a zombie, incapable of doing anything (even procrastinating), I decided to drop my prescribed morning and night 

pills just to shake things up!

I became aware of some supplements that really seemed to revitalise (at least for a time), "Happy Days: 5-HTP", so i thought I'd give them a try while I was off the tricyclic. It likely has SSRI action, which is apparently a no-no in conjunction with these.

[Image missing. Added screenshot from diary at the time in place.]

They're a night time thing, right, cos the active ingredient is the pre-cursor for Serotonin, which is then converted to Melatonin by the pineal gland when low light levels on the retina trigger the SCN to initiate a neural chain of event (gotta love how this is also revision for my neuroscience module!).

So i get a little sleep off them, as one might hope, don't particularly notice much direct improvement in getting to sleep, or much else. But over the next few days things pick up in general, and come 2am, the morning of my second exam my brain's starts banging out so many interesting idea i have to grab a pen and pad (for the second night in a row). However, after my 4 hour exam (which would have gone even better if i'd only had an extra, extra 45 minutes), the inspiration just kept coming!; i spent most of the rest of the time, until 1am, filling 10 A4 sides with idea spider diagrams, designs and research! only just managing to drag myself away for my doctor's appointment, hockey, uncharacteristically enthusiastic and incessant conversation + a quick shower, but not dinner....

Anyhow, one upshot of my, near transendant, eye opening experience and the precipitated research and synergy of ideas is knowing that:

• I've had Dysthymia for at least 7 years (just a succinct way of saying what ppl who know me already know).
• My dyslexia is more specifically Sluggish Cognitive Tempo (which again fits so well given that I've been quite literally absentmindedly wondering about my ADHD, but not so much seeing the H).
• 5-HTP (and presumably the increase in distributed Serotonin levels that it brings) fixed me!...for a few days at least. In fact it appears to have brought on 24 hours of hypomania; in this case a slightly more vivid form of the inspiration that occasionally graces me once in a blue moon.
• This may suggest that i'm also classifiable in the range of (the milder) bipolar type II, with a rating of perhaps 30/100 on the scale used in Stephen Fry's docu first part.
• I've had a very helpful little booklet, for some years, that my Mother (i believe) photocopied out of a book for me. It highly recommends 5-HTP to get sleep and well being back on track, along with a shortlist of important vitamins and trace elements that might be supplemented. So i tracked some of those down too while at the chemist, ending up with Magnesium-OK, as they just happen to have most of the important things in there (and no female specific hormones that i can see!).

Either the revitalising effects of 5-HTP wore off very fast, or the other supplements gave a helping hand, because i had a classic, rubbish day, straight after breakfast, today! Bit of a coincidence being the morning after taking the Magnesium-OK for the first time. Hoping it's transitory. I think i did need to try something more new, because the 5-HTP alone haven exactly enable much sleeping, or dreaming that i know of. And i'd rather not totally loose my mind, even in a fit of inspiration!

Another night, another lecture:

The the IET headquarters, Savoy Place, London: Dr Richard Murry (Caltec University), about his team's DARPA grand chellenge entry, 'Alice'.

[The fancy lecture theatre, with seats more comfortable than my bed and audience michrophones in every other arm rest for questions! (though it looks rubbish in my panarama)]

Alice managed 8 miles in the last challenge (won by stanford), before getting a taste for journalists and mounting a concrete barrier. Unlike Stanley, Alice appears to have been programed totally by hand, mostly by CalTec undergrads. Though i was thwarted again on the post talk question front, i did send him an email (as he suggested while leaving):

Dear Richard,
I attended the lecture this evening, but the schedule was such that you had to leave before I had a chance to speak to you.

From what I can see, Alice was/is a system entirely hand crafted by very talented engineers. Have you used, or considered using, any automated design processes (genetic algorithms, etc) or machine learning techniques to reduce the volume the programming required to produce such a complicated system?

Do you feel that a fully automated vehicle, as a finished consumer product, would need to be carefully pieced together by engineers in order that it meets safety, predictability, or any other requirements?

While having enormous respect for the abilities of you, your colleges and students, and robust development techniques, I personally can not foresee a complete solution to the real-world automated driving problem using solely these methods.

I understand you are extremely busy and these are quite open questions; a link to related discussions you have had previously would be gratefully received.

Respectfully,

Richard Lewis
Cybernetics undergraduate at the University of Reading

I'll post a reply if such unlikly event transpires...

[Alice]

...2 hours later (I shall eat my words):

Richard,
Sorry we didn't get a chance to talk at the lecture.
You are right that Alice is pretty much "hand crafted", at least atthe design stage (meaning: we use lots of optimization inside, butreal-time optimization rather than design-style optimization). Wedid do some spreadsheet analysis for looking at different sensoroptions, but I can't think of much else beyond that.

The difficult in applying automated design tools, at least for me, isfiguring out how to parametrize the system. If there are good toolsavailable for doing this for software systems (where there are almostno physical constraints that can guide the design), I would certainlybe interested in considering them. Nothing that I know of is capableof deciding what the content of a message should be and how we amodule should reactive to a directive that is sent to it. This maybe something where I am ignorant though. Definitely let me know ifyou have heard of something along these lines.

We have talked about doing work with machine learning, but haven'tcurrently put anything like that into our design. These can betricky, I think, since they can only learn about things they haveseen. For example, it seems hard to believe that we could learn howto handle the failure that occurred in our 2005 vehicle. Inretrospect there are many fixes to the problem, but with a system ofthis complexity we can't possibly explore all of the options.

Having said that, we are very interested in how to design reliablesoftware control systems. A link to a separate project that we arepursuing with those goals is
http://www.cds.caltech.edu/~murray/VaVmuri/index.php?title=Main_Page

This takes the point of view that whatever your design method, youeventually have to verify that it meets your specification and alsovalidate the design in the presence of things you haven't thoughtabout. In the project above we are focused on the first part of thisproblem.

Thanks very much for coming to the talk and for you questions!
-richard

Wireless wonderings:

Tonight's IET lecture was at Vodafone HQ, Newbury. Given by Prof. Mike Walker (big chief of R&D) on the current market, competing technologies and near-future developements in wireless telephone/data/video.

It mostly affirmed what i knew: communication and computer systems are converging: mobile phones are becoming PCs in the pocket, PCs are becoming mobile and pocket sized. Wireless services are becoming general data services (the backbone of the mob phone network is already IP packet based) hosting talk/text/browsing/video/etc.

The main limitations are radio frequency bandwidth, which ironically is most inaccesable in more developed countries where governments have tight controls on spectrum use, and are glacially slow to allocate/reallocate and expect high premiums for the privalage. This means that developing countries (like china and india) will take up future advances far more quickly, in line with their rapid economic growth. They have little legacy hardware knocking about and will be all in favour of giving out frequency bands to whoever (or even better, state controlling a single company) to ensure maximal economic benifits.

An even more fundamental limit is EM frequency: a single frequency band can only transmit roughly the number of digital data bits per second as it's carrier frequency. Current technologies are around 2-20Mhz (ish) for GPRS/3G and data rates have been prett much optimised to make use of the channels they've paid so much for (so about 5mbps currently). Using multiple adjacent frequency slots can double, triple, etc the capacity but adds much complexity (Anyone remember dual line ISDN? Thought not!). So u need higher frequencies; Do Co Mo (japan mobile company) have propose 3-4Ghz use, but this massively reduces the range of cell masts to mere 100s of meters. This means many closely spaced base stations are needed and no real coverage can be provided outside of cities and big towns...

This is a fundamental physical limitation, and just as computer perfomance increases follow the Moore's Law trend, so communications rates (mobile comms included) must follow. So they *will* be using these high frequenciesin not very long, and this *will* exclude raural residents (because there's the same infrastructure problem with physical connections over sparsely populated areas). Poor Stevie already knows this, having only *just* had ADSL become available in little ole Pailton, and even then it's not full speed of course.

The whole telecomuting thang looked (to me) set to signal the end of city life and ruaral to urban migration (mmm, GCSE geography), but i think this relatively new and increasingly powerful 'force' will consolidate cities, and perhaps aid their further growth.

Presuming an average doubling in data capaciy every year and a half (though it will undoubtably be fast, to catch up with users broadband expectations) then mobile providers will *have* to be using ~1Ghz by 2020 (though WiMax technology, etc are there *already* and is kind of in competition). From there/here it's only another 10 doublings (~15 years, though likely less) until terahertz is needed. These frequencies are closer to IR light than to microwaves, they just about pass through soft matter (like thin plastic/clothes/skin), so things like buildings would be even more of a pain in the arse than now! Then above terahertz is visible light! (maybe u could use room/street lighting to carry the data signals)

The applications for 1000'000'000'000bps wireless transmission may be elusive for now, but by 2030 they *will* exsist...any entity feeling the need to converse more thoroughly will just have to 'jack in'!


PS. The Professor made a swift exit just before i could quiz him on whether i'm likely to be able to VOIP on 3G any time soon for a semi-reasonable price! (so i can skype everywhere, and most the time for free at home) Damn, will just have to wait and see how things pan out!